The AI values change according to the standard deviation of fiber orientations in computer-generated collagen networks.

The accuracy of orientation detection was calculated by determining the average value of the maximum CDF difference of 15 computer-generated images for each standard deviation condition. The error of the algorithm was quantified by the average maximum absolute difference between the theoretical and estimated CDF values for 15 images. The error range is obtained by the standard deviation of maximum CDF difference within the 15 images. Computer generated images of increasing fiber fraction were used until errors > 10% were detected when AI < 0.94. The algorithm more accurately calculates ‘fiber’ orientation when the ‘fibers’ are less aligned. The average error of the algorithm (4.32%) in the range of AI ~ [0,0.94] leads to an average error in AI ~ 0.029. The larger errors when AI ~ [0.94,1] may be due to fiber overlap of the highly aligned fibers.
* The standard deviations of the estimated AI are at most 0.02 within 15 samples.
The AI values change according to the standard deviation of fiber orientations in computer-generated collagen networks.